Download Add Sophisticated Analytics to Your Repertoire with Data Mining

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Cluster analysis wikipedia , lookup

Multinomial logistic regression wikipedia , lookup

Transcript
Add Sophisticated Analytics to Your Repertoire
with Data Mining, Advanced Analytics and R
Why Advanced Analytics
Companies that inject big data and
analytics into their operations show
productivity rates and profitability that
are 5% to 6% higher than those of
their peers.
https://hbr.org/2012/10/making-advanced-analytics-work-for-you
http://ai.arizona.edu/mis510/other/Big%20Data%20-%20The%20Management%20Revolution.pdf
Stuck in Aspiration?
While many companies escalate to production with Advanced
Analytics, many more are stuck in an aspirational state.
MicroStrategy Bridges the Gap
Advanced Analytics Spectrum
Descriptive
Diagnostic
Predictive
Prescriptive
What Happened?
Why did it happen?
What will happen?
What should happen?
Analytical functions
Descriptive models
Predictive models
Optimization
Math, OLAP & Financial,
Operators & Statistics
Cluster & Association,
Link & Factor Analysis
Regression & Time
Series, Forecasting &
Classification
Linear Programming
Simulation
MicroStrategy Analytic Map
MicroStrategy
Functions Library
MicroStrategy Native
Data Mining Features
R Analytics
MicroStrategy Functions Library
300+ analytical functions
Statistical
Permut
FTest
HeteroscedasticTTest
HomoscedasticTTest
MeanTTest
PairedTTest
VarTest
Forecast
ForecastV
Growth
GrowthV
Intercept
Pearson
RSquare
Slope
SteYX
Trend
TrendV
Beta CDF
Binomial PDF/CDF
ChiSquare CDF/Inv
ChiSquareTest
CritBinomial
Exponential PDF/CDF
F CDF/Inv
Fisher PDF/Inv
Gamma PDF/CDF/Inv
Hypergeometric PDF
Lognormal CDF/Inv
NegativeBinomial PDF
Normal PDF/CDF/Inv
Poisson PDF
StandardNormal CDF/Inv
T CDF/Inv
Statistical Aggregate
Variance of a
Standard Deviation Population
Standard Deviation Geometric Mean
Pop
Average Deviation
Variance
Kurtosis
Skew
Weibull PDF/CDF
AvgDev
Confidence
Correlation
Covariance
Kurtosis
Skew
Standardize
Reporting
Date and Time
Math Functions
Average
Mean
Count
Sum
Maximum
Minimum
Median
Mode
Product
Rank
Percentile
“N”-Tile
N-tile by Step
N-tile by Value
N-tile by Step and
Value
Add Days
Add Months
Current Date
Current Date & Time
Current Time
Day of Month
Day of Week
Day of Year
Days Between
Month Start Date
Month End Date
Months Between
Year Start Date
Year End Date
Absolute
A-cosine
Hyp A-cos
A-sine
Hyp A-sine
A-tan
A-tan2
Hyp A-tan
Ceiling
Combine
Cosine
Hyp Cosine
Degrees
Exponent
Factorial
Floor
Integer
Ln
Log
Log10
Mod
Power
Quotient
Radians
Randbetween
Round
Sine
Hyp Sine
Square Root
Tan
Hyp Tan
Truncate
Financial
Accrued Interest
Accrued Interest Maturity
Amount Received at Maturity
Bond-equivalent Yield for T-BILL
Convert Dollar Price from Fraction to Decimal
Convert Dollar Price from Decimal to Fraction
Cumulative Interest Paid on Loan
Cumulative Principal Paid on Loan
Depreciation for each Accounting Period
Days In Coupon Period to Settlement Date
Days In Coupon Period with Settlement Date
Days from Settlement Date to Next Coupon
Double-Declining Balance Method
Discount Rate For a Security
Effective Annual Interest Rate
Fixed-Declining Balance Method
Future Value
Data Mining
Association Rules
Clustering
General Regression
Mining
Neural Network
Regression
Rule Set
Support Vector Machine
Time Series
Train Association
Train Clustering
Train Decision Tree
Train Regression
Train Time Series
Tree Model
OLAP Functions
Future Value of Initial Principal with
Compound Interest Rates
Interest Rate
Interest Payment
Internal Rate of Return
Interest Rate per Annuity
Macauley Duration
Modified Duration
Modified Internal Rate of Return
Next Coupon Date After Settlement Date
No of Coupons Settlement and Maturity Date
Nominal Annual Interest Rate
No of Investment Periods
Net Present Value
Odd Last Period / Yield
Prev Coupon Date Before Settlement Date
Price Per $100 Face Value w Odd
First Period Payment
Payment on Principal
Price
Price Discount
Price at Maturity
Present Value
Prorated Depreciation for each Period
Straight Line Depreciation
Sum-Of-Years' Digits Depreciation
T-BILL Price
T-BILL Yield
Variable Declining Balance
Yield
Yield for Discounted Security
Yield at Maturity
Running Total
Running Std Deviation
Running Std Deviation
of Population
Running Minimum
Running Maximum
Running Count
Moving Difference
Moving Maximum
Moving Minimum
Moving Average
Moving Sum
Moving Count
Moving Std Deviation
Moving Std DeviationP
First /Last in Range
Exponential Weight
Moving Avg
Exponential Weight
Running Avg
MicroStrategy Native Data Mining Capabilities
Training
Linear regressions
Ensembles
Logistic regression
Scoring
Create Dataset
Select Variables
Develop Model
Deploy Model
Detailed/Summary
Clean/ Sample
Explore/Transform
Discover Patterns
Train Model
Validate Model
Score Records
Present Results
Decision tree clustering
Time series
Score, neural networks,
rule set, SVMs
Predictive metrics are deployed like
any other metric to anyone,
anywhere, anytime
Interface Driven Predictive Model
Model Visualizations
Quality Info and Simulator
Descriptive Statistics
What is R
R is a language and environment for statistical computing and graphics
Over 5,000 packages:
Calculations and graphics
Millions of practitioners
Vibrant user communities
# 1 choice of data scientists worldwide
R Integration Pack
R is a language and environment for
statistical computing and graphics
Integrate R in 3 Steps
1
2
3
Write R script
Capture its Signature
Deploy as a metric
MicroStrategy
Desktop
MicroStrategy
Web
MicroStrategy
Mobile
MicroStrategy
Office
Competitive Advanages
Usable in any MicroStrategy Environment
One Analytic: Multiple Outputs, Multiple Applications,
Standalone Scripts
Flexible applications with
parameters/Prompts/Transaction services
Transparent R environment
Error Logging
Usable in any MicroStrategy Environment
Analytics Desktop
Mobile
Web/Visual Insight
Developer
Available on-premises or Cloud
One Analytic: Multiple Applications via Standalone Scripts
Don’t re-invent the wheel
Eg. Use the same clustering analytic to cluster
stores, employee, customers and products
Maintain and update multiple applications through a
single script
Inline scripts
Standalone scripts
• Maintenance nightmare
• Easy to Maintain
• Rife with errors
• Error Proof
• Complex for end-users
• Abstracts complexity
• Edits must be made within each calculation
• Edits made in a single location
Flexible applications with Parameters/Prompts/Transaction services
Parameter
Submit scalar values at metric execution time, e.g. (Number
of Clusters, % Confidence Interval, Train vs Score)
Prompts
Utilize all types of prompts to maximize flexibility, e.g.
(Choose a metric to forecast, choose a level at which to
aggregate)
Transaction Services
Score on-the-fly by entering in variable values on your mobile
device, e.g. (Customer comes into car dealership, enter
attribute values, and predict what car will be bought.)
Transparent R environment
Access to File System
•
•
•
•
Read/Write data files and images
Models don’t disappear into the
stratosphere
Persist R environment for further
investigation
Send insights to data scientists for
validation
Silent Installations of R Package
•
Don’t burden administrators with
package configuration
Error Logging
No Error Message Reported:
Errors Logged and Reported:
Demo
Additional resources
Try us now today
Join our Webcasts
https://rintegrationpack.codeplex.com
http://www.microstrategy.com/us/events/webcasts
Thank you.